Related Experiment Video
Updated: May 26, 2026
07:12
Using Informational Connectivity to Measure the Synchronous Emergence of fMRI Multi-voxel Information Across Time
Published on: July 1, 2014
Differentiating BOLD and non-BOLD signals in fMRI time series using multi-echo EPI
Prantik Kundu1, Souheil J Inati, Jennifer W Evans
1Section on Functional Imaging Methods, Laboratory of Brain and Cognition, National Institutes of Health, Bethesda, MD 20892, USA. kundup@mail.nih.gov
Neuroimage
|January 3, 2012
Summary
This study introduces a novel method using multi-echo fMRI to differentiate neural signals from noise. The technique effectively removes nuisance effects, improving functional connectivity analysis, especially for subcortical-cortical connections.
Area of Science:
- Neuroimaging
- Functional Magnetic Resonance Imaging (fMRI)
- Signal Processing
Background:
- Distinguishing neural signals from noise is crucial for fMRI functional connectivity studies.
- Conventional denoising methods often have limited effectiveness.
- Multi-echo fMRI allows characterization of Blood Oxygen Level Dependent (BOLD) signals based on echo-time dependence.
Purpose of the Study:
- To develop and validate a novel denoising technique for fMRI data using echo-time dependence.
- To differentiate BOLD signals from non-BOLD nuisance signals.
- To improve the accuracy of functional connectivity analysis.
Main Methods:
- Acquired whole-brain multi-echo fMRI data at three echo times (TEs).
- Applied Independent Components Analysis (ICA) to spatially concatenated data across space and TE.
- Developed summary scores to classify components as BOLD-like or non-BOLD-like based on R(2)* and S(0) changes.
Main Results:
- The TE-dependence based scores effectively differentiated BOLD-like functional network components from non-BOLD components (motion, physiological noise).
- Using non-BOLD component time courses as noise regressors significantly improved seed-based correlation mapping.
- The proposed technique demonstrated superiority over conventional methods for both individual and group analyses, particularly for subcortical-cortical connectivity.
Conclusions:
- Differentiating BOLD and non-BOLD signals based on echo-time dependence is a robust method for fMRI denoising.
- This approach enables automated removal of nuisance components, enhancing functional connectivity analysis.
- The method offers significant improvements for mapping brain connectivity, especially in challenging regions.
